Adaptation and evaluation of a weak recycling strategy for inflow boundary layers in large-eddy simulation

Abstract

This article presents a detailed analysis of an improved weak recycling method used to generate turbulent inflow conditions for large-eddy simulations. Answering some of the points raised in the literature, various flat-plate cases are considered in order to evaluate the fidelity and the robustness of the method. It is shown that the distance it takes for the canonical compressible boundary layer to forget its weak recycling inflow turbulence is rather short (< 10 δ, for Reθ up to 2 340). Yet, the study also reveals that the optimal distance between the inlet and the source plane location used to extract the fluctuations depends on the case at hand with a tendency of compressible simulations to require a sightly larger distance. Cases involving compressibility effects (M=0.8) and 3D effects (wall in translation perpendicular to the main flow) show that the recycling method remains effective in more complex flows. When using limited resolution inlet flow profiles coming from experiments as an input, spurious fluctuations are observed whose main wavelength equals the recycling distance. This study proposes a strategy to circumvent this drawback by limiting the amplitude of recycled velocity fluctuations with respect to the mean velocity at the inlet.